Predicting Chronic Opioid Use Among Patients With Osteoarthritis Using Electronic Health Record Data.

OBJECTIVE: To estimate the risk of a patient with osteoarthritis (OA) developing chronic opioid use (COU) within 1 year of a new opioid prescription by using electronic health record (EHR) data and predictive models. METHODS: We used EHR data from 13 health care organizations to identify patients with OA with an opioid prescription between March 1, 2017 and February 28, 2019 and no record of opioid use in the prior 6 months. We evaluated 4 machine learning models to estimate patients' risk of COU (≥3 prescriptions ≥84 days, maximum gap ≤60 days). We also estimated the transportability of models to organizations outside the training set. RESULTS: The cohort consisted of 33,894 patients with OA, of whom 2,925 (8.6%) developed COU within 1 year. All models demonstrated good discrimination, with the best-performing model (random forest) achieving an area under the receiver operating characteristic curve (AUC) of 0.728 (95% CI 0.711-0.745), but the simplest regression model performed nearly as well (AUC 0.717 [95% CI 0.699-0.734]). Predicted risk deciles spanned a range of 2% risk for the 10th percentile to 18% risk for the 90th percentile for well-calibrated models. Models showed highly variable discrimination across organizations (AUC 0.571-0.842). CONCLUSIONS: We found that EHR-based predictive models could estimate the risk of future COU among patients with OA to help inform care decisions. Black-box methods did not have significant advantages over more interpretable models. Care should be taken when extending all models into organizations not included in model training because of a high variability in performance across held-out organizations.

Development of a New Patient-Reported Medication Adherence Instrument: Concerns Influencing Medication Adherence.

PURPOSE: The purpose of this research was to conceptualize and develop a tool for identifying persons who are, or are likely to be, non-adherent to medications prescribed by their healthcare provider(s) by identifying concerns that patients have regarding their treatments. PATIENTS AND METHODS: The target populations were persons diagnosed with atrial fibrillation or osteoarthritis, who were prescribed anticoagulants or over-the-counter or prescription pain medications, respectively. In this two-stage, multi-year, qualitative research study, relevant concepts were explored, confirmed and refined. The focus was on non-adherence due to active (thus potentially modifiable) patient decisions to forego taking medications as prescribed. RESULTS: The most common concerns among participants with atrial fibrillation were medication-related side effects and fear of bleeding. Participants with osteoarthritis were most concerned about short-term stomach problems and long-term kidney and liver side effects. The Concerns Influencing Medication Adherence (CIMA) instrument was developed based on these concerns and those identified in the literature. It is comprised of 16 items: a core set of 11 items potentially applicable to multiple disease states, 3 items specific to atrial fibrillation, and 2 items unique to osteoarthritis. The instrument is intended to be completed electronically, and publicly available for use in direct patient care in the United States or in population health management. CONCLUSION: To our knowledge, this is the first instrument focused on medication adherence that includes documented details of patient input as recommended by the United States Food and Drug Administration guidance. Patient input is considered a key component of content validity. In this research, for example, the concerns that patients have regarding their treatments can be expected to have affected past medication adherence and can potentially impact future adherence. Although applicability outside atrial fibrillation or osteoarthritis was not assessed, the general items may be useful in assessing adherence in other chronic diseases.

Axitinib in metastatic renal cell carcinoma: patient characteristics and treatment patterns in US community oncology centers.

AIM: To study patient characteristics and treatment patterns in real-world axitinib use for metastatic renal cell carcinoma. PATIENTS & METHODS: We conducted a retrospective analysis of second- or third-line axitinib use between 1 January 2012 and 31 October 2014 in 135 metastatic renal cell carcinoma patients using the US Oncology Network database. RESULTS: Overall, 86.7% had clear cell histology, 57.8% had stage III/IV disease at diagnosis and 55.6% were poor risk by Heng criteria. Median treatment duration was 4.6 months (range: 0.03-35.49); 80.7% initiated axitinib at 5 mg/day twice daily, and 67.4% maintained this dose. Overall, 77.8% discontinued treatment, mainly due to disease progression (50.5%) and toxicity (21.9%). CONCLUSION: Axitinib usage patterns were consistent with the National Comprehensive Cancer Network Guidelines®. Ease of use among community oncologists and patient tolerance are key features of axitinib.

Health Care Costs Among Renal Cancer Patients Using Pazopanib and Sunitinib.

Publications that aim to assess the economics of different therapies are important because they complement clinical trial data and may aid in decision making. We therefore read with interest the study by Hansen et al. in the January 2015 issue of JMCP. This study compared  costs between pazopanib (PAZ) and sunitinib (SU) in the first-line treatment of patients with metastatic renal cell carcinoma (mRCC).1 The authors assessed health care costs through assignment of costs from the Truven Health MarketScan Databases to the self-reported health care resource utilization (HCRU) data from the population studied in the phase III noninferiority clinical trial COMPARZ (Pazopanib versus sunitinib in metastatic renal cell carcinoma).2 We are writing to comment on the conclusions drawn from the results presented, the methodology used, and to request additional information and clarification on data presented.